// Return a string which is the concatenation of |items|, with a new line after each line.
// "a", "b" => "a\nb\n"
private static String C(String... items) {
StringBuilder out = new StringBuilder();
for (String x : items) {
out.append(x);
out.append('\n');
}
return out.toString();
}
/**
* Assumes the given reader has only tsurgeon operations (not a tregex pattern), and returns them
* as a String, mirroring the way the strings appear in the file. This is helpful for lazy
* evaluation of the operations, as in a GUI, because you do not parse the operations on load.
* Comments are still excised.
*
* @throws IOException
*/
public static String getTsurgeonTextFromReader(BufferedReader reader) throws IOException {
StringBuilder sb = new StringBuilder();
for (String thisLine; (thisLine = reader.readLine()) != null; ) {
thisLine = removeComments(thisLine);
if (emptyLinePattern.matcher(thisLine).matches()) {
continue;
}
sb.append(thisLine);
sb.append('\n');
}
return sb.toString();
}
protected String historyToString(List history) {
String str = (String) historyToString.get(history);
if (str == null) {
StringBuilder sb = new StringBuilder();
for (int i = 0; i < history.size(); i++) {
sb.append('^');
sb.append(history.get(i));
}
str = sb.toString();
historyToString.put(history, str);
}
return str;
}
public ArrayList<String> getPairsFromSentence(String sentence) {
Collection<TypedDependency> tdl = parseSentenceTDL(sentence);
ArrayList<String> pairs = new ArrayList<String>();
for (TypedDependency td : tdl) {
StringBuilder sb = new StringBuilder();
sb.append(td.gov().originalText());
sb.append(" ");
sb.append(td.dep().originalText());
pairs.add(sb.toString());
}
return pairs;
}
protected String projectString(String str) {
if (str.indexOf('@') == -1) {
if (str.indexOf('^') == -1) {
return str;
}
return str.substring(0, str.indexOf('^'));
}
StringBuilder sb = new StringBuilder();
sb.append(str.substring(0, str.indexOf(' ')));
if (str.indexOf('^') > -1) {
// sb.append(str.substring(str.indexOf('^'),str.length()));
}
int num = -2;
for (int i = 0; i < str.length(); i++) {
if (str.charAt(i) == ' ') {
num++;
}
}
sb.append(" w " + num);
return sb.toString();
}
/**
* Assumes that we are at the beginning of a tsurgeon script file and gets the string for the
* tregex pattern leading the file
*
* @return tregex pattern string
*/
public static String getTregexPatternFromReader(BufferedReader reader) throws IOException {
StringBuilder matchString = new StringBuilder();
for (String thisLine; (thisLine = reader.readLine()) != null; ) {
if (matchString.length() > 0 && emptyLinePattern.matcher(thisLine).matches()) {
// A blank line after getting some real content (not just comments or nothing)
break;
}
Matcher m = commentPattern.matcher(thisLine);
if (m.matches()) {
// delete it
thisLine = m.replaceFirst("");
}
if (!emptyLinePattern.matcher(thisLine).matches()) {
matchString.append(thisLine);
}
}
return matchString.toString();
}
public String project(String tagStr) {
StringBuilder sb = new StringBuilder();
boolean good = true;
for (int pos = 0, len = tagStr.length(); pos < len; pos++) {
char c = tagStr.charAt(pos);
if (c == '-') {
good = true;
} else if (c == '^') {
good = false;
}
if (good) {
sb.append(c);
}
}
String ret = sb.toString();
// System.err.println("TTP mapped " + tagStr + " to " + ret);
return ret;
}
public static Tree untransformTree(Tree tree) {
TreeFactory tf = tree.treeFactory();
if (tree.isPrePreTerminal()) {
if (tree.firstChild().label().value().matches(".*_.")) {
StringBuilder word = new StringBuilder();
for (int i = 0; i < tree.children().length; i++) {
Tree child = tree.children()[i];
word.append(child.firstChild().label().value());
}
Tree newChild = tf.newLeaf(word.toString());
tree.setChildren(Collections.singletonList(newChild));
}
} else {
for (int i = 0; i < tree.children().length; i++) {
Tree child = tree.children()[i];
untransformTree(child);
}
}
return tree;
}
/**
* Prints out all matches of a tree pattern on each tree in the path. Usage: <br>
* <br>
* <code>
* java edu.stanford.nlp.trees.tregex.TregexPattern [[-TCwfosnu] [-filter] [-h <node-name>]]* pattern
* filepath </code>
*
* <p>Arguments:<br>
*
* <ul>
* <li><code>pattern</code>: the tree pattern which optionally names some set of nodes (i.e.,
* gives it the "handle") <code>=name</code> (for some arbitrary string "name")
* <li><code>filepath</code>: the path to files with trees. If this is a directory, there will
* be recursive descent and the pattern will be run on all files beneath the specified
* directory.
* </ul>
*
* <p>Options:<br>
* <li><code>-C</code> suppresses printing of matches, so only the number of matches is printed.
* <li><code>-w</code> causes the whole of a tree that matches to be printed.
* <li><code>-f</code> causes the filename to be printed.
* <li><code>-i <filename></code> causes the pattern to be matched to be read from <code>
* <filename></code> rather than the command line. Don't specify a pattern when this
* option is used.
* <li><code>-o</code> Specifies that each tree node can be reported only once as the root of a
* match (by default a node will be printed once for every <em>way</em> the pattern matches).
* <li><code>-s</code> causes trees to be printed all on one line (by default they are pretty
* printed).
* <li><code>-n</code> causes the number of the tree in which the match was found to be printed
* before every match.
* <li><code>-u</code> causes only the label of each matching node to be printed, not complete
* subtrees.
* <li><code>-t</code> causes only the yield (terminal words) of the selected node to be printed
* (or the yield of the whole tree, if the <code>-w</code> option is used).
* <li><code>-encoding <charset_encoding></code> option allows specification of character
* encoding of trees..
* <li><code>-h <node-handle></code> If a <code>-h</code> option is given, the root tree
* node will not be printed. Instead, for each <code>node-handle</code> specified, the node
* matched and given that handle will be printed. Multiple nodes can be printed by using the
* <code>-h</code> option multiple times on a single command line.
* <li><code>-hf <headfinder-class-name></code> use the specified {@link HeadFinder} class
* to determine headship relations.
* <li><code>-hfArg <string></code> pass a string argument in to the {@link HeadFinder}
* class's constructor. <code>-hfArg</code> can be used multiple times to pass in multiple
* arguments.
* <li><code>-trf <TreeReaderFactory-class-name></code> use the specified {@link
* TreeReaderFactory} class to read trees from files.
* <li><code>-v</code> print every tree that contains no matches of the specified pattern, but
* print no matches to the pattern.
* <li><code>-x</code> Instead of the matched subtree, print the matched subtree's identifying
* number as defined in <tt>tgrep2</tt>:a unique identifier for the subtree and is in the form
* s:n, where s is an integer specifying the sentence number in the corpus (starting with 1),
* and n is an integer giving the order in which the node is encountered in a depth-first
* search starting with 1 at top node in the sentence tree.
* <li><code>-extract <code> <tree-file></code> extracts the subtree s:n specified by
* <tt>code</tt> from the specified <tt>tree-file</tt>. Overrides all other behavior of
* tregex. Can't specify multiple encodings etc. yet.
* <li><code>-extractFile <code-file> <tree-file></code> extracts every subtree
* specified by the subtree codes in <tt>code-file</tt>, which must appear exactly one per
* line, from the specified <tt>tree-file</tt>. Overrides all other behavior of tregex. Can't
* specify multiple encodings etc. yet.
* <li><code>-filter</code> causes this to act as a filter, reading tree input from stdin
* <li><code>-T</code> causes all trees to be printed as processed (for debugging purposes).
* Otherwise only matching nodes are printed.
* <li><code>-macros <filename></code> filename with macro substitutions to use. file with
* tab separated lines original-tab-replacement
* </ul>
*/
public static void main(String[] args) throws IOException {
Timing.startTime();
StringBuilder treePrintFormats = new StringBuilder();
String printNonMatchingTreesOption = "-v";
String subtreeCodeOption = "-x";
String extractSubtreesOption = "-extract";
String extractSubtreesFileOption = "-extractFile";
String inputFileOption = "-i";
String headFinderOption = "-hf";
String headFinderArgOption = "-hfArg";
String trfOption = "-trf";
String headFinderClassName = null;
String[] headFinderArgs = StringUtils.EMPTY_STRING_ARRAY;
String treeReaderFactoryClassName = null;
String printHandleOption = "-h";
String markHandleOption = "-k";
String encodingOption = "-encoding";
String encoding = "UTF-8";
String macroOption = "-macros";
String macroFilename = "";
String yieldOnly = "-t";
String printAllTrees = "-T";
String quietMode = "-C";
String wholeTreeMode = "-w";
String filenameOption = "-f";
String oneMatchPerRootNodeMode = "-o";
String reportTreeNumbers = "-n";
String rootLabelOnly = "-u";
String oneLine = "-s";
Map<String, Integer> flagMap = Generics.newHashMap();
flagMap.put(extractSubtreesOption, 2);
flagMap.put(extractSubtreesFileOption, 2);
flagMap.put(subtreeCodeOption, 0);
flagMap.put(printNonMatchingTreesOption, 0);
flagMap.put(encodingOption, 1);
flagMap.put(inputFileOption, 1);
flagMap.put(printHandleOption, 1);
flagMap.put(markHandleOption, 2);
flagMap.put(headFinderOption, 1);
flagMap.put(headFinderArgOption, 1);
flagMap.put(trfOption, 1);
flagMap.put(macroOption, 1);
flagMap.put(yieldOnly, 0);
flagMap.put(quietMode, 0);
flagMap.put(wholeTreeMode, 0);
flagMap.put(printAllTrees, 0);
flagMap.put(filenameOption, 0);
flagMap.put(oneMatchPerRootNodeMode, 0);
flagMap.put(reportTreeNumbers, 0);
flagMap.put(rootLabelOnly, 0);
flagMap.put(oneLine, 0);
Map<String, String[]> argsMap = StringUtils.argsToMap(args, flagMap);
args = argsMap.get(null);
if (argsMap.containsKey(encodingOption)) {
encoding = argsMap.get(encodingOption)[0];
System.err.println("Encoding set to " + encoding);
}
PrintWriter errPW = new PrintWriter(new OutputStreamWriter(System.err, encoding), true);
if (argsMap.containsKey(extractSubtreesOption)) {
List<String> subTreeStrings =
Collections.singletonList(argsMap.get(extractSubtreesOption)[0]);
extractSubtrees(subTreeStrings, argsMap.get(extractSubtreesOption)[1]);
return;
}
if (argsMap.containsKey(extractSubtreesFileOption)) {
List<String> subTreeStrings =
Arrays.asList(
IOUtils.slurpFile(argsMap.get(extractSubtreesFileOption)[0]).split("\n|\r|\n\r"));
extractSubtrees(subTreeStrings, argsMap.get(extractSubtreesFileOption)[0]);
return;
}
if (args.length < 1) {
errPW.println(
"Usage: java edu.stanford.nlp.trees.tregex.TregexPattern [-T] [-C] [-w] [-f] [-o] [-n] [-s] [-filter] [-hf class] [-trf class] [-h handle]* pattern [filepath]");
return;
}
String matchString = args[0];
if (argsMap.containsKey(macroOption)) {
macroFilename = argsMap.get(macroOption)[0];
}
if (argsMap.containsKey(headFinderOption)) {
headFinderClassName = argsMap.get(headFinderOption)[0];
errPW.println("Using head finder " + headFinderClassName + "...");
}
if (argsMap.containsKey(headFinderArgOption)) {
headFinderArgs = argsMap.get(headFinderArgOption);
}
if (argsMap.containsKey(trfOption)) {
treeReaderFactoryClassName = argsMap.get(trfOption)[0];
errPW.println("Using tree reader factory " + treeReaderFactoryClassName + "...");
}
if (argsMap.containsKey(printAllTrees)) {
TRegexTreeVisitor.printTree = true;
}
if (argsMap.containsKey(inputFileOption)) {
String inputFile = argsMap.get(inputFileOption)[0];
matchString = IOUtils.slurpFile(inputFile, encoding);
String[] newArgs = new String[args.length + 1];
System.arraycopy(args, 0, newArgs, 1, args.length);
args = newArgs;
}
if (argsMap.containsKey(quietMode)) {
TRegexTreeVisitor.printMatches = false;
TRegexTreeVisitor.printNumMatchesToStdOut = true;
}
if (argsMap.containsKey(printNonMatchingTreesOption)) {
TRegexTreeVisitor.printNonMatchingTrees = true;
}
if (argsMap.containsKey(subtreeCodeOption)) {
TRegexTreeVisitor.printSubtreeCode = true;
TRegexTreeVisitor.printMatches = false;
}
if (argsMap.containsKey(wholeTreeMode)) {
TRegexTreeVisitor.printWholeTree = true;
}
if (argsMap.containsKey(filenameOption)) {
TRegexTreeVisitor.printFilename = true;
}
if (argsMap.containsKey(oneMatchPerRootNodeMode)) TRegexTreeVisitor.oneMatchPerRootNode = true;
if (argsMap.containsKey(reportTreeNumbers)) TRegexTreeVisitor.reportTreeNumbers = true;
if (argsMap.containsKey(rootLabelOnly)) {
treePrintFormats.append(TreePrint.rootLabelOnlyFormat).append(',');
} else if (argsMap.containsKey(oneLine)) { // display short form
treePrintFormats.append("oneline,");
} else if (argsMap.containsKey(yieldOnly)) {
treePrintFormats.append("words,");
} else {
treePrintFormats.append("penn,");
}
HeadFinder hf = new CollinsHeadFinder();
if (headFinderClassName != null) {
Class[] hfArgClasses = new Class[headFinderArgs.length];
for (int i = 0; i < hfArgClasses.length; i++) hfArgClasses[i] = String.class;
try {
hf =
(HeadFinder)
Class.forName(headFinderClassName)
.getConstructor(hfArgClasses)
.newInstance(
(Object[])
headFinderArgs); // cast to Object[] necessary to avoid varargs-related
// warning.
} catch (Exception e) {
throw new RuntimeException("Error occurred while constructing HeadFinder: " + e);
}
}
TRegexTreeVisitor.tp =
new TreePrint(treePrintFormats.toString(), new PennTreebankLanguagePack());
try {
// TreePattern p = TreePattern.compile("/^S/ > S=dt $++ '' $-- ``");
TregexPatternCompiler tpc = new TregexPatternCompiler(hf);
Macros.addAllMacros(tpc, macroFilename, encoding);
TregexPattern p = tpc.compile(matchString);
errPW.println("Pattern string:\n" + p.pattern());
errPW.println("Parsed representation:");
p.prettyPrint(errPW);
String[] handles = argsMap.get(printHandleOption);
if (argsMap.containsKey("-filter")) {
TreeReaderFactory trf = getTreeReaderFactory(treeReaderFactoryClassName);
treebank =
new MemoryTreebank(
trf, encoding); // has to be in memory since we're not storing it on disk
// read from stdin
Reader reader = new BufferedReader(new InputStreamReader(System.in, encoding));
((MemoryTreebank) treebank).load(reader);
reader.close();
} else if (args.length == 1) {
errPW.println("using default tree");
TreeReader r =
new PennTreeReader(
new StringReader(
"(VP (VP (VBZ Try) (NP (NP (DT this) (NN wine)) (CC and) (NP (DT these) (NNS snails)))) (PUNCT .))"),
new LabeledScoredTreeFactory(new StringLabelFactory()));
Tree t = r.readTree();
treebank = new MemoryTreebank();
treebank.add(t);
} else {
int last = args.length - 1;
errPW.println("Reading trees from file(s) " + args[last]);
TreeReaderFactory trf = getTreeReaderFactory(treeReaderFactoryClassName);
treebank = new DiskTreebank(trf, encoding);
treebank.loadPath(args[last], null, true);
}
TRegexTreeVisitor vis = new TRegexTreeVisitor(p, handles, encoding);
treebank.apply(vis);
Timing.endTime();
if (TRegexTreeVisitor.printMatches) {
errPW.println("There were " + vis.numMatches() + " matches in total.");
}
if (TRegexTreeVisitor.printNumMatchesToStdOut) {
System.out.println(vis.numMatches());
}
} catch (IOException e) {
e.printStackTrace();
} catch (TregexParseException e) {
errPW.println("Error parsing expression: " + args[0]);
errPW.println("Parse exception: " + e.toString());
}
}